April 25, 2026
I'm betting 303 days of my life on AI agents.
A 303-day public experiment. Build ₩20M/month (~$14K MRR) using only AI agents. 4 revenue streams. Everything open.
I've been coding for 20 years. PHP, Python, React. Solo. No team, no VC.
Then AI started doing what I do.
I panicked for a while. Then one thought changed everything:
"What if I just command the AI instead?"
That's all it took.
Starting today, I'm running a 303-day public experiment.
The bet: build ₩20,000,000/month (~$14,000 MRR) using only AI agents. 4 revenue streams. Everything public.
The 4 axes
- NaholroAI — agent-built solutions. No teaching.
- ATOZAI — agent marketplace + revenue hub (Korea)
- MrBAN Global — this site. English build-in-public.
- Investment Arena — two AIs battle in public. Paper trading now. Real money (₩10M total) starts May 30. Every decision public.
The rules
- Every number, every week. Even $0. Even negative.
- Every agent decision published. Why it bought. Why it was wrong.
- Every failure documented. Dead agents get obituaries.
Securities firms show results. I show thinking.
That's the only difference. That's the bet.
D-303 starts now.
→ Live stats: mrban.io/open
Next post: How two AIs make investment decisions — strategy reveal
— Changhyuk Ban (@mrban_ai)